SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The YadA protein of Yersinia pseudotuberculosis promotes tight adhesion and invasion into mammalian cells through β1-integrins. In this work, we demonstrate that YadA also triggers the production of interleukin-8 (IL-8) in host cells and we identify intracellular signal transduction mechanisms involved in YadA-initiated cell invasion and/or IL-8 synthesis. Tyrosine protein kinases, including the focal adhesion kinase (FAK) and c-Src, as well as the small GTPase Ras, were shown to play a significant role in both YadA-promoted cell processes. YadA-mediated cell contact led to autophosphorylation of FAK at position Tyr397 and induced GTP-loading of Ras. Furthermore, IL-8 production and invasion induced by YadA were strongly reduced in FAK- and c-Src-deficient cells and in cells overexpressing dominant interfering forms of FAK, c-Src or Ras. We also demonstrate that YadA activates the Ras-dependent Raf–MEK1/2–ERK1/2 pathway and mitogen-activated protein kinases (MAPKs) p38 and JNK. Moreover, inhibition of ERK1/2 by pharmacological agents or overexpression of dominant negative FAK, c-Src or Ras abrogated IL-8 release, whereas invasion remained unaffected. In contrast, actin polymerization and phosphatidylinositol 3-kinase (PI3K) activity is essential for YadA-promoted cell entry, but not for cytokine secretion. We conclude that YadA triggers FAK–Src complex formation and subsequent Ras activation, which leads to the stimulation of MAPKs-dependent IL-8 production or to PI3K-dependent invasion.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The enteropathogenic Yersinia species Y. enterocolitica and Y. pseudotuberculosis cause a broad range of food-borne diseases, including enteritis, enterocolitis, mesenteric lymphadenitis and autoimmune disorders (Bottone, 1997). Upon oral infection, the bacteria first encounter the intestinal mucosa and then cross the epithelial layer through M cells associated with underlying lymphoid follicles, named Peyer's patches (Marra and Isberg, 1996; Schulte et al., 2000). This process allows subsequent colonization and multiplication of the bacteria in the lymphoid follicles and the dissemination to other tissues and organs, such as liver and spleen (Grutzkau et al., 1990).

A number of enteric bacteria, including Yersinia, trigger the production of various proinflammatory substances in intestinal cells in response to bacterial invasion which attract polymorphonuclear leukocytes (PMNs) towards the location of infection (Eckmann et al., 1993; Jung et al., 1995). PMN infiltration of infected tissue mainly depends on the secretion of cytokines such as interleukin-8 (IL-8). As a consequence, multiplication of Yersinia in the Peyer's patches is accompanied by a recruitment of PMNs, but this does not seem to lead to the eradication of the bacteria (Autenrieth et al., 1996). Pathogenic Yersinia are endowed with a virulence plasmid-encoded system that delivers effector proteins, the so-called Yersinia outer proteins, Yops, into PMNs and macrophages to counteract their attack (Cornelis et al., 1998). Some of the Yop proteins, i.e. YopE and YopH, modulate the production of proinflammatory cytokines and target signalling pathways involved in the organization of the host cell cytoskeleton to inhibit bacterial phagocytosis (Juris et al., 2002; Viboud et al., 2003).

A virulence factor that stimulates the production of proinflammatory substances during Yersinia infection is the primary invasive factor invasin which seems particularly important for Yersinia to cross the intestinal epithelium in the early phase of an infection (Pepe and Miller, 1993; Marra and Isberg, 1996). This bacterial surface protein mediates cell contact and entry by binding to various members of the β1-integrin receptor family (Isberg and Leong, 1990). Invasin–integrin interaction triggers intracellular signalling pathways that involve components of the focal adhesion complex, such as Src and the focal adhesion kinase (FAK), the small GTPase Rac-1 and cytoskeletal proteins which lead to rearrangements of the actin cytoskeleton followed by bacterial phagocytosis via a zipper mechanism (Alrutz and Isberg, 1998; Isberg et al., 2000; Alrutz et al., 2001). Invasin also triggers the production of proinflammatory cytokines such as IL-8. This implicates the activation of mitogen-activated protein kinases (MAPKs), especially p38, and the transcriptional regulator NF-κB (Grassl et al., 2003a). Whether the invasin-initiated signalling pathways for cell entry and cytokine production are linked is unknown.

Besides invasin, the outer membrane protein YadA of Y. pseudotuberculosis also has the capacity to bind and invade human cells (Bliska et al., 1993; Yang and Isberg, 1993; Eitel and Dersch, 2002). YadA covers the bacterial surface by forming a capsule-like structure of lollipop-shaped surface projections, with a globular head, an intermediate stalk and a C-terminal outer membrane anchor domain (Hoiczyk et al., 2000). It confers serum resistance, mediates autoagglutination and promotes tight adherence to host cells via the extracellular matrix (ECM) proteins laminin, collagen and fibronectin (El Tahir and Skurnik, 2001). However, in mouse infection experiments only a small about fivefold difference between a Y. pseudotuberculosis yadA mutant and wild type was noticed (Han and Miller, 1997). In a previous study, we have shown that YadA-driven invasion of Y. pseudotuberculosis can be blocked with antibodies against fibronectin and β1-integrins, indicating that cell entry occurs via β1-integrins which link ECM components to the cytoskeleton (Eitel and Dersch, 2002). Little is known about the integrin-initiated cell signalling events induced by the YadA protein. We have shown that YadA-mediated uptake is blocked in the presence of protein kinase inhibitors, suggesting that these signalling molecules are important for bacterial phagocytosis (Eitel and Dersch, 2002). In this study, we demonstrate that YadA also triggers the production of IL-8 in human epithelial cells and we characterize intracellular signal transduction mechanisms induced upon YadA–host cell contact which are implicated in cell invasion and IL-8 synthesis.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

YadA of enteropathogenic Yersinia stimulates IL-8 production

A number of virulence-associated surface components of enteric bacteria, including invasin, strongly induce proinflammatory responses. To determine whether the YadA protein triggers the production of proinflammatory cytokines, we used E. coli K-12 expressing the YadA protein of Y. pseudotuberculosis Typ III (YadApstb) and Y. enterocolitica 0:8 8081v (YadAent) to infect HEp-2 cells and determined IL-8 production in the cell supernatants using an enzyme-linked immunoassay based on specific anti-IL-8 polyclonal antibodies. Comparable amounts of YadA were produced in the recombinant bacteria and in Y. pseudotuberculosis grown under YadA-inducing conditions (data not shown) and allowed us to analyse YadA-mediated IL-8 release by epithelial cells without interference by other Yersinia factors (e.g. invasin, Yops) known to affect the expression of proinflammatory cytokines. No or little IL-8 production was detected when the cells were incubated without bacteria or were treated with E. coli K-12 harbouring the empty expression vector (Fig. 1A). In contrast, strong IL-8 induction was detected with YadA-producing E. coli K-12, similar to E. coli K-12 expressing the invasin protein of Y. pseudotuberculosis (Invpstb), and elevated levels of IL-8 in the cell supernatants were retained for up to 72 h (data not shown). IL-8 production mediated by YadA was also tested in a plasmidless Y. pseudotuberculosis strain (YPIII pIB1) in the presence or absence of invasin. As shown in Fig. 1B, YadA-induced amounts of IL-8 were similar to those elicited by invasin. IL-8 production by invasin and YadA was additive and was ≈ 20-fold higher than that of the YPIII pIB1inv mutant strain.

image

Figure 1. The YadA protein of enteropathogenic Y. pseudotuberculosis and Y. enterocolitica stimulates IL-8 production. A and B. (A) E. coli K-12 strain CC118 harbouring plasmids pBAD18 (V), pRI203 (invpstb), pPD284 (yadApstb) or pJE13 (yadAent) and (B) a virulence plasmid-cured strain of Y. pseudotuberculosis YPIII pIB1 or its inv mutant derivative harbouring vector pBAD18 (V) or pPD284 (yadApstb) were applied to HEp-2 cells. IL-8 production was quantified 8 h after infection by an IL-8-specific enzyme-linked immunosorbent assay kit. Data are presented as means ± standard deviations of three independent experiments performed in duplicate. C. RT-PCR analysis of IL-8 mRNA was performed using total RNA extracted from HEp-2 cells incubated in RPMI 1640 (–) or stimulated with YadA-expressing bacteria (CC118/pPD284) for 30 min to 5 h. The GAPDH transcript served as an internal control for the amount of RNA used in each reaction.

Download figure to PowerPoint

To confirm YadA-dependent de novo synthesis of IL-8 in epithelial cells, induction of IL-8 at the mRNA level was also tested by reverse transcription polymerase chain reaction (RT-PCR) with RNA from HEp-2 cells infected with YadApstb-expressing bacteria for various times (Fig. 1C). YadApstb induced IL-8 mRNA expression and a clear IL-8 mRNA signal was first detected 1 h after infection which increased over the next 5 h. In contrast, no IL-8 mRNA was detected with control cells incubated in RPMI 1640 (Fig. 1C) and with E. coli K-12 stimulated cells at the indicated times (data not shown). This analysis of IL-8 mRNA levels together with the quantification of the IL-8 protein showed that YadA triggers the synthesis of IL-8 production.

YadApstb-mediated invasion and IL-8 production both require host tyrosine kinase activity

The YadA protein of Y. pseudotuberculosis, expressed in E. coli K-12 or in Yersinia, promotes tight adhesion and invasion into epithelial cells via β1-integrins (Eitel and Dersch, 2002). Inhibition experiments with antibodies directed against β1-integrins implied that these cell receptors are also implicated in YadA-triggered IL-8 secretion (data not shown). β1-Integrin receptors are linked to many different intracellular signalling pathways. We inquired which integrin-initiated signals contribute to YadApstb-promoted cell uptake and/or IL-8 production. One of the early cellular responses after integrin stimulation is the activation of certain protein tyrosine kinases (PTKs), associated with integrins in the cell cytoplasm (Schwartz, 2001). Fluorescence microscopy of HEp-2 cells infected with YadApstb-expressing bacteria, stained with an anti-phosphotyrosine antibody, revealed an accumulation of tyrosine-phosphorylated cellular proteins in the vicinity of cell-associated bacteria. In addition, several new tyrosine-phosphorylated proteins were identified in cell extracts early after infection, which were not seen in non-infected cells (data not shown). To further address the impact of tyrosine kinases, we incubated HEp-2 cells with increasing concentrations of general PTK inhibitors, such as genistein, herbimycin and tyrphostin before infection. We found that YadApstb-mediated cell invasion and IL-8 secretion was strongly reduced by all PTK inhibitors in a similar dose-dependent manner (Fig. 2). These results suggest that tyrosine kinases are activated early upon YadA–host cell contact and are required for cytokine expression and cell invasion.

image

Figure 2. YadA-mediated IL-8 production and invasion requires protein tyrosine kinases (PTKs) of the host cells. About 5 × 104 HEp-2 cells were pre-incubated for 30 min at 37°C in RPMI supplemented with increasing concentrations of the PTK inhibitors genistein, herbimycin and tyrphostin. Subsequently, about 5 × 106 bacteria were added to the cell monolayer and incubated at 37°C for 4 h to monitor IL-8 production (A) or for 1 h to determine the cell invasion efficiency (B). HEp-2 cells infected with E. coli harbouring the empty vector (V) or HEp-2 cells infected with YadA-expressing E. coli without the addition of PTK inhibitors (–) were used as controls. IL-8 production was quantified by an IL-8 ELISA and the number of intracellular bacteria was determined by the gentamicin protection assay. Data are presented as means ± standard deviations of three independent experiments performed in duplicate.

Download figure to PowerPoint

The focal adhesion kinase is implicated in YadApstb-mediated cell invasion and IL-8 production

The non-receptor focal adhesion protein kinase (FAK) localizes with β1-integrins and becomes autophosphorylated at position Y397 in response to many integrin-initiated signalling processes (Parsons, 2003). We first examined the capacity of YadA to stimulate phosphorylation of FAK in HEp-2 cells. Cell lysates of infected cells were analysed by blotting with an antibody specific for FAK phosphorylated at position Y397. Incubation with YadApstb-expressing bacteria resulted in an increased amount of phosphorylated FAK, which was detectable within 7 min and remained elevated for at least 30 min (Fig. 3A). A comparable amount of phosphorylated FAK was also found with an invasin-expressing E. coli K-12 strain (data not shown). To further examine the impact of FAK on cell entry and IL-8 production by YadA, we transiently overexpressed the FAK wild-type (FAK wt) protein and two FAK mutants, one of which lacks the FAK kinase activity (K454R) and the other harbours a substitution in the autophosphorylation site of FAK (Y397F) (Fig. 3B). HEp-2 cells were also transfected with the C-terminally truncated FAK derivative FRNK (FAK-related non-kinase), which does not contain the FAK kinase domain and which was shown to inhibit resident FAK when overexpressed (Hildebrand et al., 1993) (Fig. 3C). Overexpression of all FAK mutant proteins clearly reduced IL-8 production to similar levels (40–50%). Also cell invasion was decreased to ≈ 35–50% by FRNK and the Y397F mutant with respect to vector-transfected cells. The K454R variant seemed to have a less severe effect on YadA-mediated invasion. Furthermore, when FAK-deficient fibroblast cells were used in the infection assays, YadApstb-expressing bacteria barely invaded these cells and induced low amounts of IL-8, whereas efficient invasion and IL-8 secretion was observed with the same cells overexpressing the FAK wt protein (Fig. 3D). We also tested cell adhesion of the YadApstb-expressing bacteria to the different transfected cells and found comparable numbers of cell-bound bacteria (data not shown), indicating that inhibition of invasion by the FAK mutants was not caused by a reduced cell binding.

image

Figure 3. Role of the focal adhesion kinase (FAK) in YadA-mediated IL-8 synthesis and invasion. A. Activation of FAK upon YadApstb-mediated cell adhesion. HEp-2 cells were infected with YadApstb-expressing E. coli K-12 and lysed at various times (7 min to 2 h) after infection. Equal amounts of the cell extracts were loaded on 7% SDS-polyacrylamide gels and probed with anti-P-FAK Y397 or anti-FAK to confirm the presence of equal levels of FAK in the cell extracts. B. HEp-2 cells were transfected with pcDNA3.1 and constructs encoding the wild-type FAK (FAK wt) and the kinase-inactive forms of FAK, FAK Y397F and FAK K454R. (C) HEp-2 cells transfected with the vector pEGFP or pEGFP encoding the C-terminal fragment of FAK (FRNK) and (D) FAK (–/–) and FAK (–/–) + FAK wt fibroblast were infected with YadApstb-expressing E. coli. IL-8 production in the cell supernatants was determined by ELISA 4 h after infection and cell invasion was quantified 1 h after infection using the gentamicin protection assay. Ctrl: HEp-2 cells transfected with pcDNA3.1 were infected with E. coli K-12 harbouring empty vector pBAD18. The graphs show means ± standard deviations of three independent experiments performed in triplicate. Expression of the different FAK derivatives in the transfected cells were analysed by Western blotting with a monoclonal antibody directed against the human FAK protein.

Download figure to PowerPoint

c-Src plays a role in YadApstb-induced cytokine production and cell entry

Phosphorylation of FAK at position Y397 in response to integrin engagement creates a high-affinity docking site for several signalling molecules. One of which is the c-Src kinase which becomes activated upon FAK–Src complex formation (Parsons, 2003). To address the role of c-Src in YadApstb-induced cell signalling pathways, we transiently expressed a kinase-negative form of c-Src (K297M) in HEp-2 cells (Fig. 4A). IL-8 synthesis and internalization of YadA-expressing bacteria was lowered to about 40% of the vector-transfected cells, whereas bacterial adhesion to the transfected cells was not affected (data not shown). Furthermore, when Src PTK-deficient fibroblasts (SYF), lacking the Src kinases Src, Yes and Fyn, were used for infection, IL-8 secretion and cell invasion were significantly lower than with SYF cells expressing the Src protein (Fig. 4B). Notably, IL-8 production and the number of internalized bacteria recovered from SYF + c-Src cells were comparable with wild-type fibroblasts and HEp-2 cells after the same time of infection (data not shown). To corroborate these findings, we blocked the Src family kinases with increasing concentrations of the specific inhibitor PP2 (Fig. 4C). Inhibition of the Src PTKs strongly reduced IL-8 expression and impaired uptake of YadApstb-expressing bacteria in a dose-dependent manner. These results demonstrate that c-Src activity is important for YadApstb-induced cell invasion and IL-8 synthesis.

image

Figure 4. Inhibition of the Src family kinases blocks YadApstb-mediated IL-8 synthesis and invasion. (A) HEp-2 cells transfected with vector pcDNA3.1 or a construct encoding a kinase-inactive derivative of c-Src (Src K297M) and (B) SYF and SYF + c-Src fibroblasts were infected with YadApstb-expressing E. coli K-12. Ctrl: (A) HEp-2 cells transfected with pcDNA3.1 or (B) SYF cells were infected with E. coli K-12 harbouring empty vector pBAD18. C. Before infection with YadApstb-expressing E. coli K-12, HEp-2 cells were pre-treated with the indicated concentrations of the Src inhibitor PP2. HEp-2 cells infected with E. coli K-12 harbouring the vector pBAD18 (V) was used as negative control. Expression of c-Src in the cells was analysed by Western blotting with a monoclonal antibody directed against the c-Src protein. IL-8 production in the cell supernatants was determined by ELISA 4 h after addition of the bacteria, and cell invasion was quantified after 1 h using the gentamicin protection assay. The graphs show means ± standard deviations of three independent experiments performed in triplicate.

Download figure to PowerPoint

Implication of the small GTPase Ras in YadA-promoted cell uptake and IL-8 release

Previous experiments have shown that ligation to integrins also activates the small membrane-bound guanine nucleotide-binding protein Ras (Schlaepfer et al., 1994). Ras cycles between an active GTP-bound form and an inactive GDP-bound form and acts as a molecular switch linking tyrosine kinase activation to different downstream signalling events. The best-characterized downstream effector molecule of Ras is Raf-1, a serine-threonine kinase which triggers the MEK1/2–ERK1/2 MAP kinase cascade (Chang and Karin, 2001). To analyse whether YadA-triggered integrin signalling induces Ras activation, we used beads coated with the Ras binding domain (RBD) of Raf-1, which specifically binds to the GTP-bound form of Ras, to pull down and separate active GTP-bound Ras. Bead-bound Ras was subsequently detected by immunoblot analysis using a monoclonal anti-Ras antibody. As shown in Fig. 5A, activated Ras was detected in cell lysates of infected cells, whereas no or significantly lower levels of activated Ras were detected in non-infected cells or cells infected with E. coli K-12. In the next set of experiments, we investigated whether activation of Ras and the downstream effectors Raf-1 and MEK1/2 are essential for YadA-dependent IL-8 synthesis and cell invasion. For this purpose, endogenous Ras, Raf-1 and MEK1 was inhibited by coexpression of the dominant negative mutant RasN17 (Feig, 1999), dnRaf (Kolch et al., 1991) and dnMEK1 (Pages et al., 1994). In the transfected cells, expression of the mutant derivative was analysed by immunoblots, proving the reduction of phosphorylated ERK1/2 downstream in the signalling cascade (Fig. 5B). We found that the dominant negative variant of Ras significantly reduced both IL-8 induction and cell invasion to about 50% (Fig. 5C and D). In contrast, only IL-8 secretion, but not YadApstb-promoted invasion, was reduced by overexpressing dnRaf and dnMEK1 (Fig. 5C and D). Furthermore, we blocked Raf-1 activity using different concentrations of the Raf-1 inhibitor GW5074. Similarly, we found that YadA-induced IL-8 production was reduced in a dose-dependent fashion, whereas the invasion efficiency of the bacteria remained the same (data not shown). These results demonstrated that the integrity of the Ras-dependent Raf-1–MEK1/2 signalling cascade is important for IL-8 induction, but not for bacterial cell entry and this further indicated that the YadApstb-induced signalling pathways implicated in cell penetration and cytokine production might be distinct downstream of Ras.

image

Figure 5. Intracellular signalling pathways involved in YadApstb-induced IL-8 production and invasion. A. YadA-dependent activation of Ras. Uninfected HEp-2 cells (none), HEp-2 cells stimulated with GTPγ-S, or HEp-2 cells infected with E. coli K-12 pBAD18 (V) or YadApstb-expressing E. coli K-12 (YadA) were lysed, and equal amounts of soluble cellular proteins were incubated with Raf-1 RBD agarose beads. The beads were washed, boiled in sample buffer and the eluted proteins were separated on a 12% SDS-polyacrylamide gel. Western blotting was performed using an anti-Ras antibody. Recovered Ras is indicated by an arrow. B–D. Interference with the Ras–Raf–MEK1/2 pathway blocks YadA-mediated IL-8 release. HEp-2 cells were transfected with the control vector pcDNA3 and with constructs encoding dominant negative forms of Ras, Raf-1 and MEK1. (B) Cell lysates of transfected cells were analysed by immunblotting with a phosphospecific antibody to P-ERK1/2. The same membrane was stripped and re-probed with antibodies to total ERK1/2 to demonstrate the expression of the mutant signalling molecules by the inhibition of ERK1/2 activation. Transfected cells were infected with YadApstb-expressing E. coli K-12 and secreted IL-8 was determined in the supernatant by ELISA 4 h after infection (C) and intracellular bacteria were quantified by the gentamicin protection assay (D). The graphs show mean values ± standard deviations of two independent experiments performed in triplicate. Ctrl: HEp-2 cells transfected with pcDNA3 was incubated with E. coli K-12 harbouring the empty vector pBAD18.

Download figure to PowerPoint

Actin cytoskeleton rearrangements and PI3K are essential for YadA-induced invasion but are not required for IL-8 production

To test whether YadA-mediated IL-8 secretion is in fact distinct and independent of the YadA-induced cell invasion process, we blocked actin polymerization of infected HEp-2 cells by cytochalasin D. This treatment strongly inhibited bacterial phagocytosis (Fig. 6A), but did not lead to a reduction of cell adhesion (data not shown) or YadA-triggered IL-8 expression (Fig. 6B). As IL-8 production by human cells could not be induced by bacterial cell supernatants (data not shown), host cell contact but not internalization seemed essential for YadA-mediated IL-8 expression.

image

Figure 6. YadA-mediated invasion but not IL-8 production requires the host cell cytoskeleton and PI3K activity. HEp-2 cells were infected with E. coli K-12 pBAD18 (V) and YadApstb-expressing E. coli K-12 (YadApstb) for 1 h in the presence of 10 µM or 100 µM cytochalasin D (A and B) or for 1 h in the presence of different concentrations of the PI3K inhibitors LY294002 or wortmannin (C and D). Internalized bacteria were enumerated by the gentamicin protection assay (A and C) and IL-8 release was determined by an IL-8 ELISA 8 h after infection (B and D). The graphs represent mean values ± standard deviations of three independent experiments performed in duplicate.

Download figure to PowerPoint

Furthermore, we investigated whether the phosphatidylinositol-3 kinase (PI3K), which represents another target of active Ras (Rodriguez-Viciana et al., 1994), contributes to YadA-promoted IL-8 secretion and cell invasion by incubating epithelial cells with different concentrations of the PI3K inhibitors Wortmannin and LY294002 before infection. The addition of the inhibitors resulted in a significant reduction of bacterial invasion to 10–50% in a dose-dependent manner (Fig. 6C), but did not have any effect on IL-8 secretion (Fig. 6D). This confirmed our previous result, showing that IL-8 production occurs independent of the internalization of the bacteria and supported the assumption that the invasion and IL-8 signalling pathways diverge downstream of Ras.

MAP kinases are activated and required for YadA-promoted IL-8 production

The three main MAP kinase pathways involving ERK1/2, p38 and JNK were shown to play a role in the production of proinflammatory cytokines upon bacterial infection (Dahan et al., 2002; Neff et al., 2003; Wang et al., 2003). The ERK1/2 kinases are activated through the Ras–Raf–MEK1/2 signalling cascade, which was shown to be important for YadA-triggered IL-8 production (Fig. 5). Therefore, we decided to analyse the role of MAP kinase activation in YadA-triggered IL-8 expression and cell uptake in more detail. For this purpose, HEp-2 cells were pre-incubated with different inhibitors of the MAP kinase pathways. SB203580, which specifically inhibits p38 by binding in its ATP-binding pocket, PD98059 and UO126, inhibitors of MEK1/2 kinases, which activate ERK1/2, as well as SP600125, which specifically blocks JNK function, were used. As shown in Fig. 7A, both MEK1/2 inhibitors strongly reduced IL-8 expression in a dose-dependent manner. Maximal inhibition to about 10–20% was observed with 100 µM UO126 and 100 µM PD98059. p38 and JNK inhibitors only partially prevented YadApstb-induced IL-8 upregulation to about 30–50% (Fig. 7C). This suggests that the ERK1/2, but also the JNK and p38 MAP kinase pathways are involved in IL-8 expression. As shown in Fig. 7B and D, none of the inhibitors had an effect on bacterial invasion, indicating that the MAP kinases are not important for the cytoskeletal rearrangement in the bacterial phagocytosis process.

image

Figure 7. Inhibitors of MAP kinases block YadApstb-mediated IL-8 production but not invasion. YadApstb-expressing E. coli K-12 were applied to HEp-2 cells pre-incubated with different concentrations of inhibitors of the MAP kinase pathways. The MEK1/2 inhibitor PD98059 (10 µM and 100 µM) and UO126 (10 µM and 100 µM) (A and B), the p38 inhibitor SB203580 (10 µM and 100 µM) and the JNK inhibitor SP600125 (10 µM and 100 µM) (C and D) were added 60 min before the addition of the bacteria. The IL-8 contents in the supernatant of HEp-2 cells after 4 h of treatment were estimated by ELISA (A and C) and the number of internalized bacteria was determined by a gentamicin protection assay (B and D). The graphs show means ± standard deviations of three separate experiments.

Download figure to PowerPoint

To better determine the role of the MAP kinases in IL-8 upregulation, we also examined JNK, ERK and p38 activation. To do so, whole-cell extracts of infected HEp-2 cells were analysed by Western blotting using specific antibodies against the phosphorylated forms of the MAP kinases and antibodies specific for p38, ERK1/2 and JNK to control whether equal amounts of the signalling molecules are present in the lysates. We showed that ERK1/2 kinases were rapidly and strongly activated when epithelial cells were incubated with YadApstb-expressing E. coli. As shown in Fig. 8A (top), phosphorylation was maximally increased 10–20 min after infection and remained elevated over the control levels for at least 60 min after infection. Some activated ERK1/2 was also detected with cells infected with the E. coli K-12 control without YadApstb (Fig. 8A); however, the level was significantly lower compared with the YadApstb-expressing E. coli strain and decreased to background by 30 min. To further investigate whether FAK and Src are involved in the activation of the Raf–MEK1/2–ERK1/2 pathway, we analysed ERK1/2 phosphorylation in infected HEp-2 cells, overexpressing the dominant negative variants of FAK and Src. As shown in Fig. 8A (bottom), all tested mutant proteins reduced the level of P-ERK1/2 in the infected cells, indicating that YadA-mediated activaton of the MAPKs occurs via FAK and Src located upstream in the signalling cascade.

image

Figure 8. YadA-dependent MAPK activation. E. coli K-12 (CC118/pBAD18) or YadApstb-expressing E. coli K-12 (CC118/pPD284) and EGF as a control were applied to HEp-2 cells for 5, 10, 20, 30, 45 and 60 min. Whole-cell extracts were prepared and equal amounts of proteins from the extracts were used for immunoblot analysis. The activated forms of the MAP kinases ERK1/2 (A, top), JNK (B) and p38 MAP kinase (C) were detected with antibodies specific for the phosphorylated form of the MAPKs. Also HEp-2 cells transfected with vector pcDNA3.1 or constructs encoding the wild-type focal adhesion kinase (FAK wt), the kinase-inactive forms of FAK (FAK Y397F, FAK K454R), the C-terminal fragment of FAK (FRNK) or the kinase-inactive form of c-Src (K297M) were infected with YadA-expressing E. coli K-12 and the activated forms of ERK1/2 were detected in cell extracts by immunoblotting with phospho-specific ERK1/2 antibodies (A, bottom). The total amount of the MAP kinases in the cell was determined with antibodies directed against ERK1/2, p38 and JNK to ensure that equal amounts of the signalling molecules are present in the cell extracts. The different isoforms of ERK and JNK are indicated by p42/p44 and p46/p54 respectively.

Download figure to PowerPoint

A weak activation was also found for JNK and p38. Somewhat more phosphorylated JNK was detected shortly after infection and sustained for up to 60 min (Fig. 8B). The activated form of p38 slightly increased and peaked  after  60 min  of  exposure  (Fig. 8C).  No stimulation  of  p38  and  JNK  was  detected  in  cell  extracts  of control cells infected with E. coli K-12. From our results, we conclude that ERK1/2, but to a lower extent also JNK and p38 are activated by the YadA protein of Y. pseudotuberculosis.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The outer membrane protein YadA of Y. pseudotuberculosis is known to mediate tight adhesion and invasion into human cells by binding to extracellular matrix proteins (El Tahir and Skurnik, 2001). In this study, we demonstrate that YadA of enteropathogenic Yersinia also triggers the production and release of the proinflammatory mediator IL-8. In this respect, the function of YadA seems to equal that of invasin, which also induces cell entry and IL-8 synthesis by mammalian cells (Grassl et al., 2003b). Apart from invasin and YadA, many different surface components of various pathogenic bacteria were shown to trigger an increase of proinflammatory cytokines by their host cells. For instance, AfA/Dr adhesins of diffusely adhering E. coli, flagellin of enteropathogenic and enterohaemorrhagic E. coli, and outer membrane components of Salmonella enterica Serovar Typhimurium, and Vibrio cholerae all induce IL-8 production (Gewirtz et al., 2001; Dahan et al., 2002; Betis et al., 2003; Zhou et al., 2003; 2004). Among the cytokines, IL-8 is especially important because it is a potent chemoattractant for neutrophils and essential for the early host defence mechanisms to eradicate invading bacteria. In case of phagocytosis-resistant Yersinia, this response does not seem to significantly restrict the pathogen. On the contrary, PMN infiltration, which is typically accompanied by a transient separation of cell–cell contacts, was shown to facilitate access into deeper tissues and better dissemination (McCormick et al., 1997). In this regard, IL-8 production induced by YadA could support the adhesive and invasive functions of this virulence factor.

The YadA protein of Y. pseudotuberculosis promotes cell attachment by binding β1-integrin receptors (Bliska et al., 1993; Eitel and Dersch, 2002). In a previous report, we demonstrated that the interaction of YadA with the ECM molecule fibronectin not only anchors the YadA-expressing bacterium to the β1-integrins on eukaryotic cells, but also promotes efficient uptake of the microorganism by host cells (Eitel and Dersch, 2002). In the present work, the contribution of integrin-associated signalling molecules in cytokine synthesis and cytoskeletal rearrangement in response to YadA-ECM-integrin engagement was further investigated and common as well as distinct host cell factors involved in IL-8 production and bacterial invasion were identified (Fig. 9). Actin recruitment and clustering of tyrosine-phosphorylated proteins were observed underneath attached bacteria, and PTKs play a role in YadA-induced cell invasion and IL-8 release. In particular, focal adhesion kinase and c-Src, which associate to the cytoplasmic tail of clustered β1-integrins and which activate different signalling and cytoskeletal proteins, seem to be critical for both YadA-triggered processes. We showed that the expression of the truncated FAK derivative FRNK and the dominant negative mutant FAK Y397F, which prevents binding and activation of Src family PTKs (Cobb et al., 1994; Schlaepfer and Hunter, 1996), as well as the complete loss of the FAK protein, strongly inhibited YadA-promoted invasion and IL-8 production. Furthermore, pharmacological and genetic inhibition of the c-Src kinase severely diminished IL-8 secretion and bacterial uptake. The non-receptor kinases FAK and Src are usually activated by cell binding to extracellular matrix proteins such as fibronectin and their activation is an important event for integrin-mediated signal transduction pathways that coordinate cell migration, differentiation and proliferation processes (Schlaepfer and Hunter, 1998; Parsons, 2003). Notably, FAK and Src family kinases have also been shown to be required for the integrin-initiated invasion processes induced by other bacterial invasion factors, including the invasin protein of Y. pseudotuberculosis that directly interacts with β1-integrins to induce uptake in human cells, the fibronectin-binding proteins of Staphylococcus aureus and invasive factors of E. coli that mediate internalization of brain endothelial cells (Alrutz and Isberg, 1998; Reddy et al., 2000; Agerer et al., 2003). At present, much less is known about the participation of integrin-associated PTKs such as FAK and Src in proinflammatory cytokine synthesis in response to bacterial surface proteins. An α5β1-integrin binding bacterial modulin, causing host tissue pathology by inducing cytokine synthesis, termed protein I/II from oral streptococci, is known to require FAK to trigger the production and release of IL-6 and IL-8, and this protein appears to act very similar to YadA. Protein I/II-induced cytokine synthesis further involves the three main groups of MAPKs ERK1/2, p38 and JNK (Neff et al., 2003). In this study, we demonstrate that YadA-initiated integrin signalling promotes phosphorylation of the three MAPKs, especially ERK1/2, and this activation is sustained for up to 60 min. Studies with specific MAPK inhibitors demonstrated that release of IL-8 was markedly suppressed by ERK1/2 inhibitors, whereas repression by the p38 and JNK inhibitors was less pronounced, indicating that IL-8 release is mostly dependent on ERK1/2. In contrast to Listeria monocytogenes (Tang et al., 1994), MAPK stimulation is not essential for bacterial invasion, because specific inhibitors of the MAPKs only affect YadA-dependent IL-8 release but not cell entry. Inhibition of Ras, Raf-1 and MEK1 by dominant negative derivatives further demonstrated that the well-known Ras–Raf–MEK1 pathway is implicated in YadA-mediated ERK1/2 activation and IL-8 secretion. Moreover, overexpression of the interfering forms of FAK, c-Src and Ras significantly reduced YadApstb-mediated ERK1/2 activation, indicating that the activation of this MAP kinase pathway occurs through these signalling molecules. Whether low activation of p38 and JNK by YadA also depends on FAK and Src kinases is unknown. We further showed that FAK, Src and the small GTPase Ras are also required for YadApstb-promoted cell entry, whereas Raf-1, MEK1/2 and the MAPKs are dispensable. In contrast, actin polymerization and the PI3K which can be activated by FAK and Ras (Chen and Guan, 1994; Rodriguez-Viciana et al., 1994) are both essential for internalization, but are not required for IL-8 synthesis. Thus, it appears that YadA initiates a common signalling pathway for cytokine synthesis and cell entry which separates downstream of the Ras GTPase (Fig. 9).

image

Figure 9. Schematic representation of molecules involved in YadA-mediated cell invasion and IL-8 production. YadA-mediated engagement of ECM-bound integrin receptors stimulates focal adhesion kinase (FAK) autophosphorylation at Tyr 397, creating a binding site for c-Src. Recruitment and activation of Src-family tyrosine kinases can lead to the activation of Ras. Activation of the Raf-1–MEK1/2–ERK1/2 MAPK pathway is one target for the action of GTP-bound Ras, which induces IL-8 production. Ras can also activate phosphatidylinositol 3-kinase (PI3K), which might provide signals for actin cytoskeletal rearrangements required for cell invasion. CM, cytoplasma membrane; ECM, extracellular matrix.

Download figure to PowerPoint

Our findings are also consistent with other studies showing that FAK and Src play an important role in mediating signals from integrins to the MAP kinases. Most striking, adhesion of fibroblasts to fibronectin was shown to stimulate the activation of FAK and c-Src and initiates signalling events to ERK2 in a Ras-dependent manner (Schlaepfer and Hunter, 1997). In this process, phosphorylation of FAK at Tyr-397 induces the association and activation of c-Src and the formation of an active FAK–c-Src complex then leads to an increase in Ras–Raf–MEK1/2-dependent ERK1/2 activation. As overexpression of the FAK Tyr-397 derivative had a profound effect on YadA-mediated ERK1/2 activation and IL-8 synthesis, it is possible that binding of YadA to fibronectin triggers a similar signal pathway in epithelial cells. How YadA activates the Ras–ERK1/2 MAPKs cascade through FAK and c-Src is not known, but multiple signalling pathways are stimulated upon FAK–c-Src complex formation and several of these lead to Ras-dependent ERK1/2 activation (Schlaepfer and Hunter, 1998). In addition, FAK and Ras-independent mechanisms of integrin-mediated activation of Raf are known (Lin et al., 1997); however, our data indicate that they are apparently not sufficient to mediate IL-8 secretion triggered by YadA.

Similar to YadA and the protein I/II of oral streptococci, the invasin protein also induces IL-8 production through its contact with β1-integrins (Grassl et al., 2003b). Interestingly, invasin does not seem to trigger the identical set of signalling pathways. IL-8 synthesis in response to invasin occurs mainly through the activation of the p38 kinase, leading to the stabilization of IL-8 mRNA, and through JNK which stimulates IL-8 transcription, but does not seem to implicate the activation of ERK1/2 (Grassl et al., 2003a). This variation in integrin-initiated cell signalling might result from different types of β1-integrins engaged by the bacterial ligand on host cells and different ligand binding modes (i.e. direct versus indirect binding via an ECM molecule).

Another virulence plasmid-encoded factor of Y. pseudotuberculosis, YopB, induces the production of the proinflammatory cytokine IL-8 and also activates Ras, ERK and the JNK pathway (Viboud et al., 2003). Invasin and YopB-induced IL-8 production are counteracted by the virulence plasmid-encoded YopE, YopH and YopJ effector proteins, which block phagocytosis by PMNs and macrophages and of which YopE is known to inhibit the activation of JNK and ERK. As a result, only low IL-8 secretion was detected in Y. pseudotuberculosis wild-type strains harbouring the virulence plasmid. In agreement with this study, IL-8 production by invasin and YadA was detected and added up in plasmid-cured Yersinia strains (Fig. 1B), but this response was reduced when the virulence plasmid was present (J. Eitel, unpubl. data). In conclusion, YadA, invasin and the YopEHJB proteins are all implicated in the induction or suppression of signalling events leading to cytokine production and cell entry of Y. pseudotuberculosis. However, when, where and how effective these virulence factors actually contribute to these processes during the different stages of an infection is still unknown and will be a challenging task for future investigations.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Plasmids, bacterial strains, media and growth conditions

The plasmids pBAD18, pPD284 (pBAD18-yadApstb+) and pRI203 (pBR322-inv+) have been described previously (Eitel and Dersch, 2002). The plasmid pJE13 (pBAD18-yadAent+) was constructed by inserting a DNA sequence coding for yadAent into the EcoRI and XbaI sites of pBAD18. The yadAent coding sequence was amplified by PCR using primer 5′-GCGGCGGAATTCTTCATC CGGTTTGAGGTGAGG-3′ and 5′-GCGGCGTCTAGACCGA TTTCGCAGTAGATAACG-3′. E. coli K-12 (CC118) (Manoil and Beckwith, 1986), the serogroup III Y. pseudotuberculosis strain YP126 (Bolin et al., 1982) and the mutant derivative thereof YPIII pIB1 and YP31 (YPIII pIB1inv) have been described previously (Eitel and Dersch, 2002). The plasmids were introduced into the E. coli K-12 and Y. pseudotuberculosis strains by electroporation. Overnight cultures of E. coli were routinely grown at 37°C, Yersinia strains were grown at 25°C in Luria–Bertani (LB) broth. The antibiotics used for bacterial selection were as follows: ampicillin 100 µg ml−1, chloramphenicol 30 µg ml−1 and gentamicin 50 µg ml−1. To induce yadA encoded on pPD284 or pJE13, 0.2% arabinose was added to the overnight cultures and production of YadA was proven by Western blot analysis using a polyclonal antibody directed against YadA.

Eukaryotic cell lines and culture conditions

Human HEp-2 cells were cultured in RPMI 1640 media (Biochrom KG) supplemented with 5% newborn calf serum (Invitrogen) and 2 mM l-glutamine (Invitrogen) at 37°C in the presence of 5% CO2. Fibroblasts derived from FAK knock-out mouse embryos (FAK−/– cells) (Ilic et al., 1995), or from Src, Yes, Fyn triple knockout mouse embryos (SYF cells) (Klinghoffer et al., 1999), and FAK (–/–) + FAK wt and SYF + c-Src were kindly provided by Christof Hauck (Zentrum für Infektionsforschung, Würzburg, Germany) and cultured in DMEM/10% fetal calf serum supplemented with non-essential amino acids on gelatin-coated [0.1% in phosphate-buffered saline (PBS)] cell culture dishes.

RT-PCR analysis

Total RNA of 4 × 105 infected HEp-2 cells seeded in six-well plates was extracted using the EasyRNA Kit (Qiagen) and 1 µg of the total RNA was reverse transcribed with the Moloney murine leukaemia virus (M-MLV) reverse transcriptase (Promega) according to the manufacturer's instructions. Subsequently, the reverse transcriptase was inactivated by heating the sample at 70°C for 15 min and cDNA products were amplified by PCR in 35 cycles in 50 µl of samples of a mixture containing 0.3 µl of the Taq DNA polymerase (5 U ml−1), 5 µl of the cDNA and 35 µl of the prepared master mix including the IL-8 and GAPDH specific primers of the Human IL-8 and GAPDH Genes Dual PCR Kit (Biozol). The PCR temperatures used were 94°C for 30 s (denaturing), 55°C for 30 s (annealing) and 72°C for 2 min (polymerization) followed by an extension of 5 min at 72°C. PCR products were separated on 1% agarose gels and stained with ethidium bromide.

Il-8 ELISA

The amount of IL-8 produced and released by HEp-2 cells into the culture supernatant was determined by enzyme-linked immunoabsorbent assay (ELISA) with an IL-8 Assay Kit (Trinova) as described in the manufacturer's instruction. One hundred microlitres of the supernatant of the infected cells and a series of twofold dilutions of the standard were mixed with 50 µl of a diluted (1:1000) biotinylated mouse anti-human monoclonal antibody directed against IL-8. The mixture was added to the 96-well coated plate and incubated for 1 h at room temperature. After the plate was washed three times, 100 µl of a 1:2000 dilution of the horseradish peroxidase (HRP) conjugated antibody was added to each well. The reaction was developed with tetramethylbenzidine (TMB)-substrate reagent and was stopped with 2 N H2SO4. Optical density at 405 nm (OD405) was measured with a Microplate Reader 680 (Bio-Rad). Recombinant IL-8 was used as a standard.

Preparation of cell extracts, gel electrophoresis and Western blotting

HEp-2 cells in six-well plates were serum-starved for 20 h and infected with 3 × 107 bacteria. At various times after infection, the cells were washed with PBS and resuspended in 200 µl of lysis buffer (1% Triton X-100, 20 mM Tris-HCl, pH 8.0, 100 mM NaCl) or 200 µl of SDS sample buffer. Extracts were boiled for 5 min and 15 µl of the samples were separated on 7–15% SDS-polyacrylamide gels. Subsequently, the proteins were transferred to immobilon membranes (Millipore) and incubated with TBSTM (Tris pH 7.5, 500 mM NaCl, 0.1% Tween and 5% skim milk powder) for at least 1 h to block non-specific protein binding. Primary antibodies against ERK1/2, phospho-ERK1/2, p38, phospho-p38, JNK and phospho-JNK (New England Biolabs), anti-P-tyrosine PY99 (Santa Cruz), c-Src (Calbiochem), FAK (Santa Cruz), P-FAK Y397 (Calbiochem) were diluted 1:1000 in TBST (Tris pH 7.5, 500 mM NaCl, 0.1% Tween) and applied to the membrane for at least 1 h at room temperature or overnight at 4°C. After washing, the blots were incubated with HRP-conjugated anti-mouse IgG (diluted 1:5000) in TBSTM. Immunoreactive bands were visualized by enhanced chemiluminescence using the Western Lightning Chemiluminescence Reagent (Perkin Elmer).

Cell infection for IL-8 determination

Before infection, HEp-2 cells were serum-starved for 20 h. For cell infection, overnight cultures of the bacteria were collected by centrifugation and washed three times in PBS pH 7.5. The optical density of the culture was determined and appropriate dilutions of the bacteria were taken for cell infection in a bacterium-to-cell ratio of 100:1. HEp-2 cells were washed three times with PBS and incubated in RPMI supplemented with 20 mM Hepes (pH 7.0) and 0.4% bovine serum albumin (BSA) before the addition of bacteria. After infection, the cells were incubated at 37°C for 2 h. Subsequently the cultures were washed and further incubated for at least 4 h in the presence of gentamicin to kill extracellular bacteria. The culture supernatant was removed and centrifuged for 10 min to pellet residual bacteria and cell debris and was then taken for IL-8 detection.

Cell adhesion and invasion assay

In preparation of the cell adhesion and uptake assay, 5 × 104 HEp-2 cells were seeded and grown overnight in individual wells of 24-well culture plates (Biochrom). Cell monolayers were washed three times with PBS and incubated in RPMI 1640 medium supplemented with 20 mM Hepes (pH 7.0) and 0.4% BSA before the addition of bacteria. Overnight cultures of the bacteria were washed once in PBS and the optical density of the culture was determined. Subsequently, appropriate dilutions of the bacteria were taken for cell infection in a bacterium-to-cell ratio of 100:1. The number of living bacteria in the culture was also determined by plating of 1:10 serial dilutions on LB medium. Approximately 5 × 106 bacteria were added to the monolayer and incubated at 20°C to test for cell binding or at 37°C to test for invasion. One hour post infection, the cells were washed extensively with PBS. The total number of adherent bacteria was determined by cell lysis using 0.1% Triton and plating on bacterial media. Bacterial uptake was assessed 60 min after infection as the percentage of bacteria which survived killing by the addition of the antibiotic gentamicin to the external medium, as described previously (Dersch and Isberg, 1999). For each strain, the level of bacterial adhesion and uptake was determined by calculating the number of colony-forming units relative to the total number of bacteria introduced onto monolayers. The experiments were routinely performed in triplicate.

Ras activation assay

Ras activation in infected cells was determined with the Ras activation assay kit (Chemicon International) as suggested by the manufacturer. Briefly, HEp-2 cells were grown to 70–80% confluency and serum-starved overnight. Subsequently, cells were incubated in 20 mM Hepes (pH 7.0) and 0.4% BSA, or infected with E. coli K-12 and YadApstb-expressing E. coli K-12 in 20 mM Hepes (pH 7.0) and 0.4% BSA medium for 1 h. Uninfected and infected HEp-2 cells were washed three times in PBS, resuspended in the assay buffer of the kit and detached from dishes with a cell scraper. For a positive control, a portion of the uninfected cell lysate was mixed with GTPγ-S for 15 min at 30°C. Cell lysates (treated with bacteria, GTPγ-S or untreated) were mixed with the Raf-1 RBD agarose slurry (50%) for 1 h at 4°C. Finally, the beads were collected by centrifugation, washed three times with the assay buffer and resuspended in SDS sample buffer. Activated Ras in the cell lysates was visualized by Western blot using an anti-Ras monoclonal antibody, followed by a secondary antibody conjugated to HRP and Western Lightning Chemiluminescence reagent (Perkin Elmer).

Inhibitors of IL-8 production and cell invasion

For the inhibitor studies, increasing concentrations (0–100 µM) of tyrosine kinase inhibitors genistein (Sigma), herbimycin (Sigma) and tyrphostin AG 1478 (Sigma), MEK1/2 inhibitors U0126 (New England Biolabs) and PD98059 (Sigma), p38 inhibitor SB203580 (Sigma), JNK inhibitor SP600125 (Sigma), Src family inhibitor PP2 (Calbiochem), PI3K inhibitors wortmannin (Sigma) and LY294002 (Sigma), Raf inhibitor GW5074 (Calbiochem) and inhibitor of actin polymerization cytochalasin D (Sigma) were diluted in RPMI 1640 medium supplemented with 20 mM Hepes (pH 7.0) and 0.4% BSA, added to human cells and incubated for 1 h at 37°C before infection with bacteria.

Transient transfection

Expression constructs encoding FAK wt, kinase-inactive FAK Y397F and FAK K454R, FRNK and kinase-inactive c-Src K297M were kindly provided by Christof Hauck. Dominant negative derivatives of Ras, Raf-1 and MEK1 were obtained from Eberhardt Hildt. HEp-2 cells seeded in 24-well plates for 24 h were transfected with the empty vectors pcDNA3.1 (Invitrogen), pEGFP (Clontech) or the respective constructs (250–800 ng). Coexpression of a GFP reporter construct was used to determine the transfection efficiency. For transfection, Lipofectamine 2000 (Invitrogen) was used, according to the manufacturer's instruction. Transient infections were performed in triplicate. The efficiency of transfection was proven by the analysis of GFP expression and the synthesis of the different signalling molecules was proven by Western blot. Cells with a transfection rate of about 70% were obtained and used for the infection with bacteria at a multiplicity of infection (moi) of 100 about 48 h after transfection.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We thank Drs Martin Fenner, Astrid Lewin and Eckhard Strauch for helpful discussions and critical reading of the manuscript. We particularly thank Drs Christof Hauck and Eberhardt Hildt for expression plasmids. This work was supported by Grant DE616/2 from the Deutsche Forschungsgemeinschaft.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References